The cerebral cortex controls mammalian behavior essentially through processing sensory input to then coordinate efferent activity via projection to subcortical structures. Anatomically, it is composed of a complex six-layer laminar structure. Sensory information processing occurs in the more superficial layers; cortical projection neurons (CPNs) arise in the deep layers and project axons either to other parts of the cortex or to subcortical areas, thus representing the most important cellular modulators of cortical output. Reestablishment of a CPN's complex, specific connectivity would therefore be an important goal of any repair strategy for cortical injuries. Recently, there has been much interest in using cellular therapy to replace neurons, usually through implantation of embryonic stem cells (ES cells). Although these cells can be readily induced to differentiate into neurons, it remains to be established whether there are qualitative differences between various differentiating strategies and types of neurons seen after transplantation. Our own studies assessing two pools of in vitro conditioned neuronal ES cells revealed an unexpected qualitative difference in their capacity to form CPN's after transplantation into neonatal cortex. Thus, while one cell pool produced virtually no CPN's, the other produced a robust subcortical projection that was remarkably anatomically appropriate. Furthermore, by serial comparisons of these two pools of cells, we were able to establish that this behavior was driven by a relatively small subset of transcription factors that function during development to produce the CPN population. These findings therefore raise the possibility that one can specifically modulate the neuronal phenotype of a cell repair treatment prior to transplantation through induction (or inhibition) of finite numbers of genes in vitro. In this R21 proposal, we propose to optimize the production of this specific neuronal cell subset through enriching the cell subpopulation for markers associated with CPN production. Additionally, we will assess the necessity and sufficiency of the identified genes in this process through either inducing their activity in the non- CPN producing pool of cells or silencing them in the CPN producing pool. With the results obtained, we will then be in a better position to assess whether this strategy has the capability of effecting functional improvement, thus moving it closer to eventual utility in the clinic, primarily as a treatment of cortical birth injuries. Brain injuries represent a significant public heath problem that is not only quality of life diminishing but also extremely costly. Improving recovery from these devastating problems is therefore of great relevance. Embryonic stem cells represent possibly the best potential cell type for use in brain transplantation. The ability to control the neuronal differentiation of these cells so that they can produce specific cell subsets should therefore be of great value in developing better therapeutic strategies for this problem. [unreadable] [unreadable]